Your child finishes a Junior Maths Challenge past paper under the clock. The room marks it and says something like "52 out of 135". Then it offers one button: Take the wrong ones to the clinic. What it does not say — ever, on any paper — is whether 52 would have been a Gold, a Silver, a Bronze or nothing at all. That silence is deliberate, it is the single decision in this room we are most often asked about, and there are two separate reasons for it.
Reason one: for most papers, there is nothing to compare against
This one is boring and completely decisive.
UKMT sets its certificate bands after the fact, against how that year's cohort actually performed. Those boundaries are not published alongside most past papers. When we loaded the twenty-four papers that sit in the room — twelve Junior, twelve Intermediate, spanning 2015 to 2026 — twenty-three of them came with no grade boundaries at all. Exactly one does.
So for twenty-three papers out of twenty-four, a certificate verdict would not be a lookup. It would be an invention. We would be picking a number, dressing it in the language of an awarding body, and handing it to a thirteen-year-old.
Here is what the room does instead, and the detail matters more than it sounds:
Where boundaries are absent, the mark bar draws no tick row. Not a greyed-out one. Not an approximated one. Not a faint dotted line with a "roughly" next to it.
A greyed row would still be a claim. It would say there is a band here and we are being coy about it, and a child would read the position of the bar against it anyway. An approximated row would be worse: it would be a guess wearing the costume of a fact. So when we do not know, the row is not drawn. The absence is visible as absence.
Reason two: we would refuse even if the boundaries existed
Suppose UKMT published boundaries for all twenty-four papers tomorrow. We still would not do it, and this is where the honest thing to say is that it is a rule about the whole product, not a judgement about this one room.
Nowhere in aitutors.me will you find a number that guesses at a future result — no figure claiming to say how ready a child is, no position relative to other children, no forecast of what they will get on the day. Not on the exam surfaces, not in the reports, not in the tutor's mouth. There is a test in the codebase that sweeps every exam surface looking for this language and fails the build if it finds any. It is one of the few things we have automated a conscience for.
The reason is not squeamishness. It is that these numbers are almost always worse than the raw information they replace, in two specific ways.
They collapse a rich thing into a thin one. A mark of 52 out of 135 sits alongside 25 answered questions, a set of wrong ones, and the named mistakes underneath them. A single verdict throws all of that away and leaves a label. The label feels more informative and contains strictly less.
They cannot be acted on. Ask what a child should do differently on Tuesday because a paper "would have been a Bronze". There is no answer. The verdict does not point anywhere. It is a mood, not an instruction.
What a number a child can act on looks like
Compare the two things side by side.
| A verdict | A mark, plus the clinic |
|---|---|
| "That would have been a Silver." | "52 out of 135." |
| Nothing follows from it | Seven wrong answers, filed by name |
| Feels like a judgement of the child | Is a description of one hour |
| Cannot be wrong out loud, so cannot be checked | Recomputed from the answers, every time |
| Points nowhere | Points at Take the wrong ones to the clinic |
That last row is the whole argument. The clinic is where every wrong answer lands, filed not under the paper it came from but under the mistake behind it, with a count next to it. A child looking at "3 things to fix" has something to do this week. A child looking at a certificate band has something to feel.
We would rather they had something to do. That is not a softer position than giving a verdict — in practice it is a much harder one, because it means the product has to actually be able to name what went wrong. We wrote about how that filing works in practice organised by the mistake.
What we will happily record: the thing that actually happened
None of this makes us squeamish about certificates. Quite the opposite.
When your child sits the real challenge through school and a result comes back, you can record it. In the parent dashboard under exam prep, a past exam carries a result, and for UKMT the options are exactly the real ones: Gold, Silver, Bronze, Took part, Through to the next round — plus an optional typed score, grade or level if you have one. Your child can record it too.
That is a record of what happened. It is grounded in an actual event, awarded by an actual body, on an actual day. It cannot be wrong in the way a forecast can be wrong, because it is not making a claim about the future at all. Keep it, be pleased about it, look back at it next year.
The distinction the whole room turns on is this: we will hold the results your family earned, and we will not manufacture the ones they have not.
What this means for you as a parent
- Read the mark, then read the clinic. The mark tells you where the hour landed. The clinic tells you what to do about it. Only one of those is a plan.
- Do not go hunting for boundaries to fill the gap. The numbers floating about for old papers are mostly reconstructions. Comparing a practice mark taken at a kitchen table against a cohort boundary from 2017 is a comparison of two things that do not match.
- Watch the mark on the same paper family over a term, not against a band. Rising is rising. That is the signal, and it costs nothing to read.
- Record the real one when it comes. It matters, and it is the kind of thing a fourteen-year-old likes to be able to look up.
If a product ever tells you exactly what you were hoping to hear about a child's future performance, the useful question is where that confidence came from. In this room the honest answer would have been "we made it up", so we do not say it.
FAQ
Why won't the room tell me whether my child's mark would have earned a certificate?
Two reasons, both true. The practical one: grade boundaries are simply not published for most past papers — of the twenty-four papers loaded in the room, twenty-three have no boundaries at all. The principled one: we do not put numbers that guess at a future result anywhere in the product, and we were not going to make an exception here.
What does the room show instead?
The mark. "{marks} out of {max}", then a button that says "Take the wrong ones to the clinic". Where a paper has no published boundaries, the mark bar draws no tick row at all — not a greyed-out one, not an approximated one. Nothing is implied that we cannot stand behind.
Can we record a real certificate my child actually won?
Yes. In the parent dashboard, past exams can carry a recorded result — for UKMT that is Gold, Silver, Bronze, Took part, or Through to the next round, plus an optional typed score or grade. That is a record of what happened on the day. It is a different kind of thing from a guess made in advance, and it is the only kind we keep.
Related reading
- Sit a Past Paper Under the Clock
- Practice Organised by the Mistake
- What Parents See While a Child Practises
- The Junior Maths Challenge: Dates, Format and How to Get Ready
Duke Harewood built aitutors.me for his own KS3-aged daughter. She has asked more than once what a paper "would have been", and the answer is still no. Updated 10 September 2026.